Aqueous fluids rich in CO2 and NaCl play an important role in the geochemical evolution of mafic and ultramafic rocks in the upper mantle and lower crust, where they act as strong agents of metasomatism and are capable of transporting significant mass and modifying bulkrock composition. Keys to understanding fluid-rock interaction in these settings are constraints on the solubilities of silica and magnesium in mixed fluids at high pressure and temperature. This dissertation outlines three experimental projects which provide new data and improve thermodynamic modeling on mineral dissolution in the system MgO-SiO2-H2O-CO2-NaCl. The first investigation focused on the solubility of magnesite (MgCO3) in H2O-NaCl-CO2 solutions at 10 kbar, from 600-800 C, at a mole fraction of CO2 (XCO2) of 0.05 and XNaCl from 0 to halite saturation. These experiments show that the solubility of MgO derived from magnesite in H2O-NaCl-CO2 fluids is significantly lower than that of Ca derived from calcite (CaCO3). The measurements improve thermodynamic modeling to allow prediction of magnesite or dolomite CaMg(CO3)2 solubility in H2O-CO2-NaCl fluids at high pressure and temperature. The second study investigated the solubility of quartz in H2O-CO2 fluids at 500-800 C and 10 kbar. My results show that, at constant P and T, quartz solubility declines with increasing XCO2 at all temperatures investigated. Critical to quantifying silica dissolution is understating its interactions with the H2O component of any mixed crustal fluid. This can be modeled by accounting for the hydrous components of dissolved silica via the equilibrium

〖SiO〗_(2,s)+nH_2 O=Si(OH)_4∙(n-2)H_2 O_((aq)) quartz hydrated monomer

where n is the “hydration number”, which corresponds to the total number of H2O molecules consumed by the transfer of SiO2 from quartz to the fluid. Previous studies argue that n is a constant independent of P and T. My results require that n varies with P and T, and I show that the variation can be linked to changes in the dielectric constant of H2O. The final study investigated the concentration of dissolved SiO2 in H2O-CO2 fluids in equilibrium with quartz + magnesite, talc + magnesite, and forsterite + magnesite at 500-800 �C and 10 kbar. Results show that total Si concentration in equilibrium with quartz + magnesite is identical to that in equilibrium with quartz alone at the same P, T, and XCO2. In contrast to changes in quartz solubility, Si concentration buffered by talc + magnesite and forsterite + magnesite decrease with rising temperature. Comparison to thermodynamic models reveals that the models are generally accurate, provided that silica hydration is taken into account, though I find that a revised model based on Newton and Manning (2002, 2008) yields better agreement with experiment than that of the Deep Earth Water model (Sverjensky et al., 2014; Fang and Sverjensky 2019). I show how my results can be incorporated into models of Si metasomatism during infiltration of H2O-CO2 fluids, as in the mantle wedge above subduction zones. Taken together, the experimental and modeling results will provide new insights into SiO2 and MgO mobility in H2O-CO2-NaCl fluids in deep crustal and upper mantle settings.

Brines are potentially important agents of mass transfer in the lower crust and subduction zones. Experiments and models typically assume that alkali-halide solutions can be approximated by the system NaCl-H2O. However, KCl is also likely to be abundant in high pressure and temperature fluids, and the effect of this solute is unknown. We experimentally determined the solubility of quartz in H2O-KCl solutions at 7.5 and 10 kbar and 600-800°C. We used hydrothermal piston cylinder methods and determined solubility by weight loss.

At 10 kbar and 800°C, quartz solubility is 1.23 mol SiO2/kg H2O in pure H2O. Quartz solubility increases slightly with addition of a small amount of KCl up to 0.06 bulk KCl mole fraction (XKCl). At higher XKCl, quartz solubility decreases exponentially to 0.19 mol SiO2/kg H2O at sylvite saturation (XKCl = 0.70). Quartz solubility decreases with T at constant XKCl. At 700°C and 600°C, quartz solubility at sylvite saturation (XKCl = 0.60 and 0.54, respectively) is 0.22 and 0.17 mol SiO2/kg H2O. At 7.5 kbar, “salting in” or silica solubility enhancement is observed at XKCl < 0.025. Solubility then decreases monotonically.

Solubility variations at different P, T and salt types can be compared using the relative solubility (i.e., observed solubility divided by that in pure H2O at P, T). In H2O-NaCl solutions at 10 kbar, relative solubility decreases with rising NaCl mole fraction independent of temperature. This is not seen at high P in KCl-H2O solutions. Relative quartz solubility measurements in KCl-H2O solutions at 7.5 and 10 kbar are higher than in NaCl-H2O solutions at any given salt mole fraction. Values in KCl-H2O solutions are comparable to those in NaCl-H2O solutions at lower P of ~5 kbar. This could be a result of the formation of hydrated KCl-SiO2 complexes.

Textural and geochemical analyses of many petrologic examples indicate mass transport of silicates and mineral paragenesis enabled by alkali-rich, low-a(H2O) fluids. This newly discovered enhancement of silica solubility in KCl-H2O solutions implies the possibility of significant silica metasomatism involving potassic salt solutions in deeper environments.

Understanding the interactions between calcite and aqueous solutions is important when studying the deep carbon cycle. The present study investigates the solubility of calcite in salt-H₂O fluids at 700 °C and 8 kbar. The investigated salts included NaCl, KCl, LiCl, and CsCl. All experiments were conducted in a piston cylinder apparatus. The results show that calcite solubility increases with increasing concentration of any individual salt. The data were successfully fit to simple functions of the square of salt mole fraction. At a given salt concentration, the solubility enhancement increases with decreasing salt cation size. Experiments in the fluids with mixtures of multiples of salts were used to derive simple relations that can be used to predict calcite solubility in a wide range of salt solutions at the studied conditions. The results provide a basis for extending this approach to different pressures, temperatures, and salt compositions.

This dissertation consists of a series of studies investigating geochemical and petrological aspects of mantle minerals, namely, olivine, spinel, orthopyroxene, and clinopyroxene. The first study is an experimental look at the solubility of the clinopyroxene mineral diopside (CaMgSi₂O₆) in natural fluids. Results allow insight into the chemistry of high P-T fluids and the behavior of a major mantle mineral in equilibrium with such fluids. The second study focuses on iron isotopic compositions of mantle minerals as powerful tracers for geochemical processes in the mantle, such as partial melting, metasomatism, and oxidation. To address this, we studied inter-mineral iron isotopic fractionation of minerals from five distinct mantle-xenolith lithologies from San Carlos, Arizona. We compared the fractionations with opposing calculations predicting equilibrium iron fractionation at high temperatures, and applied the results to implications of petrogenesis of the xenoliths. The last study is an experimental determination of equilibrium magnesium isotope fractionation between spinel, forsterite, and magnesite from 600°C to 800°C. In these experiments we implemented the three-isotope method with forsterite and magnesite, and with spinel and magnesite, at three different temperatures in high-pressure piston cylinder apparatus for varying lengths of time, using carbonate as the exchange medium. The result is the first rigorous high temperature experimental calibration of magnesium isotope fractionation of mantle minerals, and is generally consistent with expectations based on crystal chemical environment of Mg in these phases. The combination of experimental petrology with isotope geochemistry is a powerful approach for understanding mantle processes. Comparisons of these types studies with natural samples and theoretical predictions provide new insights into Earth's mantle.

The mineral apatite incorporates OH, F, and Cl into its structure, so volatile concentrations in apatite could be used as a proxy for their abundances in magmas. I investigated this by conducting experiments on F-OH exchange between apatite and a simplified arc basalt at 10 kb and 1100 to 1250 °C for 48 hours. The experiments succeeded in growing, for the first time, apatites that were large enough (25-50 µm) for analysis by secondary ion mass spectrometry (SIMS), measurements were also conducted with the electron probe microanalyzer (EPMA). Results indicated that F is preferentially partitioned into apatite, whereas OH slightly favors the liquid. The exchange of F and OH between apatite and melt can be represented as Fliquid + OHapatite = OHliquid + Fapatite, for which the equilibrium constant is Keq=[Fap][OHliq]/[OHap][Fliq], assuming unit activity coefficients. Weighted least squares fitting to data yielded Keq = 59.